Evidence from more than a dozen prospective studies supports the growing conviction that alcohol advertising influences adolescent alcohol use. Yet these studies have deficiencies that limit what can be learned from them. They focus on a narrow range of exposures via traditional mass media, often rely solely on retrospective self-reports of exposure, and do not adequately capture the psychological processes that are hypothesized to moderate and mediate the relationship. Recent advances in the area of ecological momentary assessment (EMA) provide the opportunity to address these gaps and substantially advance the field's understanding of the effects of alcohol advertising in all of its forms. EMA solicits data from respondents at the time of exposure and in real world contexts, providing detailed, sensitive and ecologically valid assessment of the micro-processes involved in ad response. Such data are crucial for validly testing mediating and moderating processes. This study will use EMA as one component of an intensive longitudinal design in which a racially and ethnically diverse cohort of 700 youth will be followed for two years. These youth will be assessed at repeated time points (0, 6, 12, 18, and 24 months) through paper surveys and 14-day time/event sampling using EMA via handheld computers. Two types of interviews will be programmed on the handheld computers: random prompts and exposure-related event recordings. Event recordings will be initiated by participants each time they see an alcohol ad and will ask participants about the specifics of the exposure. Event reports will also instigate a short survey assessing participants' alcohol-related thoughts at that moment. Random prompts will be initiated by the device 3 times per day and include the same survey, thus providing a no exposure control condition against which to compare post-event reports. Paper surveys will be completed at the start of each 14-day EMA sampling to assess changes in drinking over the past 6 months. This design will provide a detailed assessment of youths' exposure to a wide variety of media types that would be impossible to measure with retrospective surveys, and will permit evaluation of the following theory-driven hypotheses: (1) youth will have more favorable drinking expectancies and believe drinking is more common among their peers after exposure to alcohol ads than when they are not exposed; (2) youth with drinking experience or intent to drink will be more affected by exposure to ads than youth without experience or intent; (3) exposure via social media will have greater impact than exposure via other channels; (4) ads eliciting greater identification with drinkers, more positive feelings, and less skepticism toward the advertiser or message will have greater impact than ads eliciting less identification and positive affect and more skepticism; and (5) exposure-related shifts in alcohol- related cognitions, as measured by EMA, will mediate changes in drinking intentions and actual drinking over time. Results of this study will have important implications for the design of alcohol use prevention programs and the ongoing policy debate on whether to restrict alcohol advertising and portrayals in the media.